- \n
- Placement season is no longer a safety net: 38% of IIT students who registered for 2024 placements went unplaced \u2014 double the 2022 rate.<\/li>\n\n\n\n
- New graduates have fallen to 7% of Big Tech hires, and AI has eliminated the “boilerplate junior” role \u2014 not the junior who can direct AI.<\/li>\n\n\n\n
- The market now pays for verifiable proof of work: real commit history, a contest rating, deployed projects, and demonstrated AI-steering skill.<\/li>\n\n\n\n
- Second year is the highest-leverage time to start: interns convert to full-time offers at 63%, and early proof compounds for two full years before placements.<\/li>\n<\/ul>\n\n\n\n
The numbers no placement brochure will show you<\/h2>\n\n\n\n
RTI disclosures covering all 23 IITs show that of the 21,500 students who registered for 2024 placements, roughly 8,000 \u2014 38% \u2014 were still unplaced<\/a> at the end of the season, double the 19% of 2022. At newer IITs the unplaced share touched 40%<\/a>.<\/p>\n\n\n\n
The global picture is harsher. SignalFire’s State of Tech Talent report<\/a> found that new graduates now make up just 7% of Big Tech hires, down from 15% before the pandemic \u2014 and new-grad hiring overall has fallen by half. In the same research, 37% of hiring managers said they would rather use AI than hire a fresh graduate.<\/p>\n\n\n\n
Read those two paragraphs together and the conclusion writes itself: a Tier-1 admit is no longer a job guarantee. It’s an entry ticket to a tournament that now starts two years before placement season.<\/p>\n\n\n\n
What actually changed: AI ate the “trainable junior”<\/h2>\n\n\n\n
Companies used to hire juniors to write boilerplate \u2014 CRUD endpoints, form validation, glue code. AI tools now produce that instantly, which is why Stack Overflow’s analysis of the junior market<\/a> shows the traditional entry path narrowing so sharply.<\/p>\n\n\n\n
But look at what job descriptions now ask for: “experience with AI-assisted development workflows” has moved from a bonus line in 2023 to a baseline requirement. The junior who is hired in 2026 isn’t paid to type code. They’re paid to direct<\/em> AI, review its output, catch what it misses, and ship the result. That is a learnable, provable skill \u2014 and second year is exactly when to build it.<\/p>\n\n\n\n
Your competition is already public<\/h2>\n\n\n\n
GitHub’s Octoverse 2025 report<\/a> shows India added over 5.2 million developers last year \u2014 the largest source of new developers in the world \u2014 and became the largest open-source contributor base on the platform, ahead of the US. GitHub projects 57.5 million Indian developers by 2030.<\/p>\n\n\n\n
Here’s what that means practically: the strongest students in your batch are already building in public. Their commit graphs, contest ratings and deployed projects are visible to any recruiter with a browser. A blank GitHub profile in third year isn’t neutral \u2014 next to theirs, it’s a signal.<\/p>\n\n\n\n
![\"proof-of-work-ladder\"](\"https:\/\/tierones.io\/blog\/wp-content\/uploads\/2026\/07\/proof-of-work-ladder-opt-1024x576.jpg\")
<\/figure>\n\n\n\nThe second-year playbook<\/h2>\n\n\n\n
Five moves, in order of leverage. None of them require permission, money, or a placement cell.<\/p>\n\n\n\n
- \n
- Ship two real projects.<\/strong> Not tutorials \u2014 deployed things with a URL, used by at least one person who isn’t you. A mess bill-splitter your hostel actually uses beats a cloned Netflix UI every single time.<\/li>\n\n\n\n
- Make your GitHub tell a story.<\/strong> Recruiters read commit history, not star counts. Steady weekly commits on real problems signal more than a one-week hackathon spike. India is now GitHub’s largest open-source contributor base \u2014 reviewers know what a genuine graph looks like.<\/li>\n\n\n\n
- Hold one verifiable rating.<\/strong> LeetCode, Codeforces \u2014 pick one and climb honestly. Ratings are the only skill signal that can’t be embellished on a r\u00e9sum\u00e9, which is precisely why serious employers trust them.<\/li>\n\n\n\n
- Learn to steer AI, on the record.<\/strong> Build with Copilot or Claude, but review every line, write commits that explain the why, and be ready to defend any file in an interview. “I shipped this with AI and can explain all of it” is the strongest sentence a 2026 candidate can say.<\/li>\n\n\n\n
- Do one real internship before third year.<\/strong> NACE’s 2026 data<\/a> shows 63% of interns convert to full-time offers \u2014 the highest in five years \u2014 and intern hires are 32% more likely to be retained after one year<\/a> than hires without internship experience. One real internship in second year compounds into everything after it.<\/li>\n<\/ol>\n\n\n\n
Why second year specifically<\/h2>\n\n\n\n
Because it’s the only year where the math is entirely in your favour. You have enough fundamentals to build, and two full years for proof to compound before anyone else in your batch starts trying. The student who starts shipping in second year doesn’t compete in placement season \u2014 they’ve usually been hired before it begins.<\/p>\n\n\n\n
That’s the entire premise Tierones is built on: verified 2nd-year developers from IITs, NITs and IIITs, matched to startups that hire on proof \u2014 real commits, real ratings, real shipped work<\/a> \u2014 not on a r\u00e9sum\u00e9 PDF.<\/p>\n\n\n\n
FAQ<\/h2>\n\n\n\n
When should I start applying for internships?<\/h3>\n\n\n\n
By the middle of second year. The conversion and retention data above is the reason: internships are the highest-probability path to a full-time offer, and the earlier the first one lands, the longer the compounding runs.<\/p>\n\n\n\n
Do startups actually hire 2nd-years?<\/h3>\n\n\n\n
Yes \u2014 because startups hire on demonstrated skill, not graduation year. What they need is evidence they can verify in minutes: a real commit history, a rating, something deployed.<\/p>\n\n\n\n
Is a high CGPA still worth anything?<\/h3>\n\n\n\n
It clears filters; it doesn’t win offers. Treat CGPA as table stakes and spend the surplus hours on proof of work \u2014 the data above shows which one the market is actually paying for.<\/p>\n\n\n\n
- Make your GitHub tell a story.<\/strong> Recruiters read commit history, not star counts. Steady weekly commits on real problems signal more than a one-week hackathon spike. India is now GitHub’s largest open-source contributor base \u2014 reviewers know what a genuine graph looks like.<\/li>\n\n\n\n
- Ship two real projects.<\/strong> Not tutorials \u2014 deployed things with a URL, used by at least one person who isn’t you. A mess bill-splitter your hostel actually uses beats a cloned Netflix UI every single time.<\/li>\n\n\n\n